Commentary

Big Data Transformed Madison Avenue, Now Agencies Are Betting On Data To Transform Themselves

Madison Avenue has been transformed by the emergence of Big Data, now it’s betting big on data to transform itself. Publicis’ $4 billion acquisition of Epsilon, on the heels of Interpublic’s $2 billion acquisition of Acxiom, and Dentsu Aegis Network’s $1.5 billion acquisition of Merkle, comes as another big agency holding company, WPP, is expected to shed part of its investment in the kind of insights that used to fuel advertising and media -- primary research -- vis a vis a partial sale of Kantar later this year.

The restructuring of the holding companies mirrors an overall shift away from old school insights created via research, human judgement and gut, toward one that is data-driven, algorithmic and, ultimately one that leans heavily on machine logic.

As author Ken Auletta so eloquently characterized it in his extraordinary examination in “Frenemies,” it is the shift from “Mad Men to Math Men.”

It also comes as conventional agencies face more competitive threats from big consultants than ever before (note Accenture’s acquisition of Droga 5).

Interestingly, the Big Data restructuring of Madison Avenue, makes one of the big holding companies stand out even more for its neutral approach -- Omnicom -- which is betting big on people-based insights vs. data-based ones.

2 comments about "Big Data Transformed Madison Avenue, Now Agencies Are Betting On Data To Transform Themselves".
Check to receive email when comments are posted.
  1. andy brown from Consultant, April 15, 2019 at 11:51 a.m.

    I think you are right Joe, there is a swing of the pendulum towards “big data” just now in agency media planning and execution. As Kantar, the company that first brought set top box measurement to the US (and globally) to exist alongside peoplemeter panels for TV, the value is not lost on us. There is no question deterministic big data sources have a role to play in targeting and will in certain cases remove the need or limit the use for certain probabilistic solutions. However, no big data set ever told a user WHY a consumer did something!. Very few big datasets can have the accuracy to overcome the bias of capture. In some cases they are new walled gardens, with limited visibility over the wall. Netflix claim not to need ratings to make decisions. Imagine a world where Netflix is losing growth or going backawards in subscription. The first question Mr Hastings will ask, is “where did my audience go?” The no doubt very valuable first party data generated from the Netflix platform will not give context to determine where the audience has gone. It is the same with advertiser first party solutions, they can use it to drive greater use amongst their existing franchise, but how do they conquer from their competitors or people new to the category?

    The solution, is as always in life, is balance (ie. the pendulum will swing back). Using large scale deterministic data in combination with high quality probabilistic to provide genuine, stable high quality insights to clients is the role of the media/market researcher of today. Maybe things don’t move as fast as Dave Morgan would like (linking to another story), but as I am often quoted as saying “almost everthing in media is measurable, we just have to overcome economics and politics” ! 

  2. Ed Papazian from Media Dynamics Inc, April 15, 2019 at 5:08 p.m.

    Good points, Andy. When will people realize that data---big or otherwise--- must be projectable to whatever total universe that is being considered and open to realistic evaluation relative to common sense and---gasp!----human judgement based on practical experience, consideration of likely trade-offs, etc.?

Next story loading loading..